I am an independent Quantitative Researcher specializing in the intersection of Financial Econometrics and Deep Sequence Learning. With an academic foundation in Financial Management and advanced postgraduate coursework in Data Science (University of Verona), I build high-performance, GPU-accelerated pipelines for financial time series forecasting, volatility modeling, and risk-aware backtesting.

My research focuses on developing robust, leakage-free pipelines that can extract predictive structures from highly noisy, non-stationary intraday financial data (XAUUSD, Forex, and Crypto).

• GPU-Accelerated Data Engineering: Utilizing CuPy to bypass CPU bottlenecks, enabling high-speed rolling calculations (volatility, technical indicators, and custom market regimes) across millions of high-frequency data rows.
• Deep Sequence Modeling: Implementing state-of-the-art architectures in PyTorch—including TimesNet, TCN, Transformers, and recurrent models (LSTM/GRU)—to capture multi-periodicity and long-range temporal dependencies in financial series.
• Causal Risk-Aware Backtesting: Architecting fractal-based causal Stop-Loss/Take-Profit labeling engines and dynamic ATR-based position-sizing simulators to enforce realistic trading constraints without look-ahead bias.

Note: My comprehensive 50-page quantitative research monograph outlining these methodologies on a 12-year 5-minute XAUUSD dataset is available upon request.

I am actively looking for fully funded PhD positions in Quantitative Finance, Computational Economics, or Financial Econometrics where I can apply and expand my research in deep learning pipelines. I am also highly open to Quantitative Researcher / Developer roles in FinTech or DeFi.

• 📧 Email: mohammadreza.akhlaghi@studenti.univr.it (or your personal email)
• 💼 LinkedIn: Mohammadreza Akhlaghi
• 📁 Featured Research: See my pinned repositories below.